Neo

Today we unveiled Neo Residency, a new program for startups and high-agency student teams. 🎉 We’re replacing our best-known program, Neo Accelerator, with something even better and more selective. 🧵

conviction almost killed our company. for 2 years, we built tech that no one else had ever attempted. but the smartest thing we ever did was walk away from it. my co-founder, sahaj, and i had spent years building a wearable device that could read your brain. the idea: control your phone or computer without saying a word out loud. no awkward "hey siri" in public. just think it, and it happens. the technology was working. but when we finally connected it to siri, alexa, and chatgpt to test it out, we had to face a hard truth: none of us actually wanted to use our own product. walking away from two years of work is one of the hardest things i've done as a founder. but here's what that moment actually taught me: the best product decisions come from humility, not conviction. you can spend all the time you want designing something perfect. but the most valuable thing you can do is ship a quick experiment, feel it yourself, and be willing to acknowledge the hard truths. that's how we found wispr flow. not through a grand vision. but from walking into our own office one day and seeing half the team talking to their computers through $10 mics and realizing: the behavior change was already happening. we just had to get out of the way and build around it. if you're building something and wrestling with when to stay the course versus when to change direction, i think you'll find something useful in this one. link to the full podcast in comments!

I just came across an insane stat: 60% of office workers have an injury that nobody talks about. Take a look at your hands right now. Chances are they're slightly curled, even when resting. Maybe your wrists ache. Your neck tilts forward without you noticing. This isn't normal. It's what 8 hours a day at a keyboard does to you. Repetitive Strain Injury affects 60% of office workers. Carpal tunnel cases doubled in the last decade. The average knowledge worker types 50,000-80,000 keystrokes per day. Your job is literally breaking your body. Yet we're all acting like this is just the cost of doing business. Here's what we're all missing: You can't ergonomic-accessory your way out of fundamentally broken input. The problem isn't your posture. It's that we're forcing humans to spend 2,000+ hours per year in a position our bodies were never designed for. Hunt-and-pecking through emails. Switching between keyboard and mouse 200 times an hour. Craning forward because the screen is always too far or too close. Meanwhile, we have the technology to change this. What if you could write that email while pacing? Draft that doc while stretching? Respond to Slack without ever touching a keyboard? That's why we built Wispr Flow. Not basic speech-to-text that makes you edit for 10 minutes. Actually clean, structured output. In any app. Without your hands. You talk like a human. It outputs like you typed it perfectly. We spent years building the AI that understands context and intent. Not just transcribing words, but capturing what you actually mean. Because knowledge work shouldn't require physical sacrifice. Your brain does the thinking. Your voice does the input. Your body stays intact.

The rule of law applies to everyone - including state governments. The Arizona Attorney General’s charges are baseless and a clear overreach. It’s gamesmanship from a politician who’s up for reelection. The charges claim that putting money on “a contingent future event or occurrence” is illegal. If they can bring these criminal charges against Kalshi, they could do the same to traditional derivatives on CME and Nasdaq, from options trading, to interest rate swaps and grain futures. This is exactly why federal law is not optional: states cannot override it for political maneuvers or special interests. We filed suit in federal court against Arizona last week to address this directly. Instead of letting the federal court evaluate the case on the merits, the AG is attempting to subvert the appropriate judicial process by making an end run with sham charges in state court. If this were about consumer protection, the state would focus on predatory practices of the gambling industry’s addiction-driven business model. Instead, a regulated exchange is being targeted to protect incumbents and prevent consumers from choosing. But we have close to 400k customers in Arizona, about 5% of the state’s population and growing, and we will fight for them. We will not be intimidated and we will keep building.

90 minutes left to attempt generational wealth. The odds are not zero. And it's free. You owe it to your grandchildren.

I ❤️ Tarek & Luana. 🤩 Great interview.

This is an insightful and provocative post by Wispr Flow cofounder @SahajGarg6. sahajgarg.github.io/blog/cognitive…

if you're not putting ur future cofounder into your shitty notion crm three years before meeting them again wyd

i'm sure we're all familiar with the quiet and lonely moments of company building... never underestimate the power of a little dog like roy. stayed with me at the office all day and walked home in the rain after :)

This is the era of the one-person marketing team. Launch dozens of high-performing landing pages from Claude and Clay. Using @tryflint’s MCP and API.

Check my new blogpost on AI Agents here: dev.to/dimimil/licens…

glad we broke the protocol and drank some brazilian beers. Brahma > Guinness?

listenedto.today i documented a song of the day every day for the last year. see what im listening to

This manifesto introduces Agent-Native Environments (ANE), a framework for designing development environments that enable reliable agent execution. ANE systems emphasize environmental determinism, structural consistency, and rapid validation to allow agents to produce production-grade changes with minimal human oversight. In such environments, engineering productivity scales primarily with system design, architectural taste, and the ability to construct environments that agents can reliably operate within. marcelochaman.ca/repository/age…

had a few interesting conversations recently! im curious what if subagents didn't know they're subagents? the standard subagent has isolated context, handed a summary and returns findings. imo this works great for narrow tasks but for harder tasks, the summary is probably lossy. the parent spent thousands of tokens building up intuition about implicit constraints and dead ends, and we compress all of that into a paragraph. the subagent will often have to read the files again to get the full context with its cold start. i always notice my claude code usage increase at a much higher rate when it uses subagents instead, maybe we fork the conversation. the child gets the parent's full prefix (already computed via KV cache, basically free) but don't include the tool call that spawned it. from the child's perspective, the conversation just naturally pivoted to a new focus the orchestrator knows about the fork. the model doesn't. when we join back, we just attach the child's findings/output back to the parent this is basically just fork() with copy-on-write. after branching, the child appends its own suffix and the join is still text-level. @sgl_project SGLang already supports fork/join abstractions + we have things like prefix caching, RadixAttention this might just like not work at all... has this already been done? is the token consumption/latency of subagents with a cold start studied? my intuition tells me it's probably a hybrid

The team behind Megatron Core MoE is incredible -- it was such a joy to get to work on this with @EthanHe_42 at NVIDIA. Some especially fun pieces of the work: - NVFP4 training recipes - Parallel folding - Deep EP arxiv.org/abs/2603.07685

There are moments where you stop and look back at the journey — and realize just how rare it is. Today is one of those moments. I’m incredibly proud to share that @forethought_ai has signed a definitive agreement to be acquired by @Zendesk .

😙 @neo scholar retreat!!

Truly so much fun last night. Great vibes. Thanks to @johnamqdang who took it away representing team @adaption_ai after I lost my voice. 🔥

We’ve raised $400M at a $9B valuation. Investors include Georgian, G Squared, Prysm, 1789, YC, Coatue, a16z, Craft, and QIA, with strategic investments from Accenture, Databricks, Okta, and Tether. We’re also lucky to have incredible individuals backing us, including Shaq and Jared Leto. This funding will help us scale our ambition and expand beyond coding into AI systems that center human creativity. Replit is now used at 85% of the Fortune 500. We have an opportunity to help shape the future of work. One where AI abstracts away the boring parts and humans shine as creative directors. We’re also investing more globally, particularly in Europe, Asia, and the Middle East. Innovation can come from anywhere in the world, and we want to help unlock it.